Learn
Every word and every method, in plain language.
The explanations on this page are stored with the data they describe, so what you read here is exactly what the pages calculate.
How we test a signal
1. Define it first
A signal is written down as a rule before it is tested, and the rule is not adjusted afterwards to make the result look better.
2. Use only what was known
On each past day the test uses only figures that were public on that day - a result filed in May is not used in April.
3. Count honest periods
Results over periods that share days are not counted twice. An edge needs a t-value of 2 or more over at least 8 periods that stand apart.
4. Publish everything
Tests that found nothing are shown beside the ones that found an edge. Track record
Look a word up
Words we use · 13
- Edge
- A test result strong enough to count: a t-value of 2 or more over at least 8 periods that do not overlap.
- Weak, No edge, Against
- Results short of an edge. Weak: pointing the expected way, but not strongly enough. No edge: nothing to see. Against: clearly the opposite of what was expected.
- t-value
- The average result divided by how much it varies from period to period. The larger it is, the less likely the result is luck.
- Periods that do not overlap
- Two one-month results that share days are not two pieces of evidence. We count only periods that stand apart.
- Top third against bottom third
- Each period the shares are ranked on a signal. We compare what the highest third earned with the lowest.
- Standing
- A company's rank among the index members of the day. 1 is the best of the day, not a perfect company.
- Quality checks
- Ten fixed yes/no standards, the same for every company on every day.
- Group
- Where a share or sector stands in a ranking: highest, upper, middle, lower or lowest group of the day.
- India VIX
- The volatility the option market expects for NIFTY 50 over the next month, as a yearly figure.
- Put-call ratio
- Open put contracts for each open call contract.
- Max pain
- The index level at which option buyers of the nearest monthly expiry would lose the most. It describes positions; it is not a forecast.
- Breadth
- How many of the 50 shares take part in a move: shares up and down, and the share of them above their own average price.
- Delivery
- The part of a day's traded shares that was really delivered, not bought and sold again within the day.
The quality checks · 14
- Return on equity
- 15 % or more - for every company
- Return on equity, 5-year average
- 15 % or more - for every company
- Return on capital employed
- 15 % or more - for companies other than financial ones
- Return on assets
- 1 % or more - for financial companies
- Revenue growth a year, 5 years
- 10 % or more - for every company
- Profit growth a year, 5 years
- 10 % or more - for every company
- Profit against a year ago
- higher - for every company
- Debt to equity
- 1 or less - for companies other than financial ones
- Interest cover
- 3 times or more - for companies other than financial ones
- Cash from operations to profit
- 0.8 or more - for companies other than financial ones
- Gross bad loans
- 3 % or less - for financial companies
- Net bad loans
- 1 % or less - for financial companies
- Gross bad loans against a year ago
- not higher - for financial companies
- Promoter shares pledged
- 10 % or less - for every company
A company's mark is the number of its checks passed. A check without a figure is not counted.
Scorecard measures: Profitability · 11
- Return on equity
- Profit of the last four quarters as a share of shareholders' funds.
- Return on capital employed
- Profit before interest and tax of the last four quarters against equity plus debt.
- Operating margin
- Operating profit of the last four quarters as a share of revenue.
- Net margin against 3 years ago
- Net margin of the last four quarters minus the same three years earlier.
- Return on equity, 5-year average
- Average of the return on equity over the last five years (at least four).
- Return on equity, weakest point in 5 years
- The lowest four-quarter return on equity of the last five years: how bad a bad year was.
- Return on capital employed, 5-year average
- Average of the return on capital employed over the last five years (at least four).
- Net margin against 5 years ago
- Net margin of the last four quarters minus the same five years earlier.
- Return on assets
- Profit of the last four quarters as a share of total assets.
- Operating profit to assets
- Profit before provisions of the last four quarters as a share of total assets.
- Return on assets, 5-year average
- Average of the return on assets over the last five years (at least four).
Scorecard measures: Growth · 7
- Revenue growth, 1 year
- Revenue of the last four quarters against the four before.
- Profit growth, 1 year
- Profit of the last four quarters against the four before.
- Revenue growth a year, 3 years
- Yearly growth rate of four-quarter revenue over three years.
- Profit growth a year, 3 years
- Yearly growth rate of four-quarter profit over three years.
- Revenue growth a year, 5 years
- Yearly growth rate of four-quarter revenue over five years.
- Profit growth a year, 5 years
- Yearly growth rate of four-quarter profit over five years.
- Steadiness of growth
- Share of the last eight quarters in which revenue and profit grew on the year.
Scorecard measures: Health · 10
- Debt to equity
- Borrowings against shareholders' funds; lower is safer.
- Debt to equity against 5 years ago
- Debt to equity now minus five years earlier; a fall is better.
- Interest cover
- Profit before interest and tax against the interest paid, last four quarters.
- Cash conversion
- Operating cash flow against profit over the same months.
- Free cash flow margin
- Operating cash flow minus capital spending of the last full year, as a share of revenue.
- Piotroski score (8 tests)
- Eight yes/no tests of profit, cash, debt, share count, margin and efficiency against a year ago.
- Gross bad loans
- Gross non-performing assets as % of advances; lower is better.
- Net bad loans
- Net non-performing assets as % of advances; lower is better.
- Gross bad loans against a year ago
- Change of the gross bad-loan ratio over four quarters, in percentage points.
- Provisions against operating profit
- Share of the last four quarters' operating profit set aside as provisions.
Scorecard measures: Valuation · 5
- Price to earnings
- Price against the earnings per share of the last four quarters; lower is cheaper.
- Price to book
- Price against shareholders' funds per share; lower is cheaper.
- P/E against its own 5 years
- Where today's P/E stands among its own values of the last five years (0-100).
- P/B against its own 5 years
- Where today's P/B stands among its own values of the last five years (0-100).
- Dividend yield
- Dividends of the last 12 months as a share of the price.
Scorecard measures: Ownership · 4
- Promoter shares pledged
- Pledged or encumbered promoter shares as % of all shares; lower is better.
- Promoter holding against a year ago
- Change of the promoters' holding over four quarters, in percentage points.
- Institutions, last quarter
- Change of the holding of foreign and domestic institutions over one quarter.
- Institutions, 1 year
- Change of the holding of foreign and domestic institutions over four quarters.
Scorecard measures: Management · 2
- Tone of the last results
- Average tone of the last four results filings: positive 1, neutral 0.5, cautious 0.
- One-off items in profit
- Exceptional items of the last four quarters against profit before tax; lower is cleaner.
Technical terms: Flow · 5
- Delivery spike
- The share of traded stock taken for delivery is 2.5 standard deviations above its 20-day average. Read as investors, not day traders, behind the day's trade.
- Long build-up, three days running
- Price and futures open interest both rise for three days. Read as fresh long positions being added.
- Short build-up, three days running
- Price falls and futures open interest rises for three days. Read as fresh short positions being added.
- FII index futures unusually long
- The long share of foreign institutions in index futures enters the top 10 % of its history. Read as crowded optimism; often a contrarian caution.
- FII index futures unusually short
- The long share of foreign institutions in index futures enters the bottom 10 % of its history. Read as crowded pessimism; often a contrarian positive.
Technical terms: Momentum · 10
- MACD crosses above its signal line
- The MACD line crosses above its 9-day signal line. Read as momentum turning up; more reliable in the direction of the main trend.
- MACD crosses below its signal line
- The MACD line crosses below its 9-day signal line. Read as momentum turning down.
- RSI leaves the oversold zone
- RSI 14 rises back above 30. Read as selling pressure easing after a sharp fall.
- RSI falls into the oversold zone
- RSI 14 falls below 30. Read by contrarians as stretched to the downside; in a strong down-trend it can stay there.
- RSI rises into the overbought zone
- RSI 14 rises above 70. Read as stretched to the upside; in a strong up-trend it is a sign of strength instead.
- RSI leaves the overbought zone
- RSI 14 falls back below 70. Read as buying pressure easing after a sharp rise.
- Momentum rising
- MACD histogram positive and larger than three days ago. MACD histogram positive and larger than three days ago.
- Momentum fading
- MACD histogram positive but smaller than three days ago. MACD histogram positive but smaller than three days ago.
- Momentum recovering
- MACD histogram negative but less so than three days ago. MACD histogram negative but less so than three days ago.
- Momentum falling
- MACD histogram negative and more so than three days ago. MACD histogram negative and more so than three days ago.
Technical terms: Price · 4
- New 52-week high
- The highest close in a year, the first in 20 trading days. Read as strength: there are no trapped buyers above.
- New 52-week low
- The lowest close in a year, the first in 20 trading days. Read as weakness; contrarians watch for exhaustion.
- Gap up of 3 % or more
- The open is at least 3 % above the previous close. Read as strong overnight news; the question is whether the gap holds.
- Gap down of 3 % or more
- The open is at least 3 % below the previous close. Read as bad overnight news; sharp gaps sometimes reverse.
Technical terms: Trend · 11
- Golden cross
- The 50-day average crosses above the 200-day average. Read as the start of a long up-trend. It is slow: much of the rise is already behind.
- Death cross
- The 50-day average crosses below the 200-day average. Read as the start of a long down-trend. Often arrives after most of the fall.
- Close back above the 200-day average
- The close crosses above its 200-day average. Read as the long-term trend turning up.
- Close below the 200-day average
- The close crosses below its 200-day average. Read as the long-term trend turning down.
- Up-trend gains strength
- ADX rises above 25 while +DI is above -DI. Read as a trend, not a range: pullbacks tend to be bought.
- Down-trend gains strength
- ADX rises above 25 while -DI is above +DI. Read as a trend, not a range: rallies tend to be sold.
- Strong up-trend
- Close above the 50-day average, which is above the 200-day average, and ADX above 25. Close above the 50-day average, which is above the 200-day average, and ADX above 25.
- Up-trend
- Close above both the 50-day and the 200-day average. Close above both the 50-day and the 200-day average.
- Sideways
- Close between its 50-day and 200-day averages: no clear trend. Close between its 50-day and 200-day averages: no clear trend.
- Down-trend
- Close below both the 50-day and the 200-day average. Close below both the 50-day and the 200-day average.
- Strong down-trend
- Close below the 50-day average, which is below the 200-day average, and ADX above 25. Close below the 50-day average, which is below the 200-day average, and ADX above 25.
Technical terms: Volatility · 2
- Breakout up from a squeeze
- The close moves above the upper Bollinger band within 5 days of an unusually narrow band. Read as the start of a move after a quiet spell.
- Breakdown from a squeeze
- The close moves below the lower Bollinger band within 5 days of an unusually narrow band. Read as the start of a fall after a quiet spell.
Technical terms: Volume · 2
- Volume spike on an up day
- Volume is 2.5 times its 20-day average and the close is up. Read as large buyers at work.
- Volume spike on a down day
- Volume is 2.5 times its 20-day average and the close is down. Read as large sellers at work, or as a selling climax.
Descriptive analytics only. Never investment advice.